We've found that durable workflows is a much needed primitive for agents control flow. They give a structure for deterministic replays, observability, and, of course, fault tolerance, that operators need to make the agent loop reliable.

Isn't that what they call "Harness engineering"?

I agree and I think a really wonderful way to encode agentic control flow would be with Polynomial Functors.

https://arxiv.org/abs/2312.00990

we work on this issue in healthcare (genhealth.ai) where it's imperative to get every step correct. not easy. a valuable solution at the intersection of browser, code, lmms. there r far more layers of browser interaction than just imgs and dom.

I always wonder with these posts: - are they talking about coding (where I am the control flow) - or RPA agents (in which it is obvious) ? - also don't use llm for deterministic tasks

We built https://aflock.ai/ (open source) to help with this. Constraining activity tends to work well

> "Agents need control flow, not more prompts"

Can't wait for ya'll to come full circle and invent programming from first principles.

Eventually we'll all come to the inevitable conclusion that for a task to be fully automated there should be neither human nor genie in the loop.

This exactly why I’m building aiki to be a control layer for harness execution. I don’t think the model companies will ever give us the neutral layer we need.

9 different frameworks being pushed in the comments of this thread. 2026 truly is the year of agents.